Human cytomegalovirus (HCMV) is an important pathogen which causes severe morbidity in immunocompromised individuals, including newborn infants. In particular, infants born with congenital CMV infection are at high risk for poor neurodevelopmental outcome. Since pre-existing maternal immunity protects against severe disease caused by HCMV, there is considerable interest in developing vaccines designed to prevent the substantial morbidity associated with congenital infection. Vaccines for CMV have been difficult to evaluate in the preclinical setting, however, because of the strict species specificity of cytomegaloviruses. Since the guinea pig cytomegalovirus (GPCMV) crosses the placenta, causing infection in utero, the guinea pig provides a model system in which to test vaccines for the prevention of CMV disease. Unfortunately, this system has not to date been exploited to its fullest potential, largely because so little is known about the molecular biology of GPCMV. However, recently the GPCMV genes encoding the major humoral immune target glycoprotein B (gB) and the cell-mediated immune target pp65 (UL83) have been cloned, characterized and expressed. Subunit vaccine studies are thus now feasible in this model. Therefore, these studies propose three major aims. First, the immunogenicity and protective efficacy of a recombinant form of envelope glycoprotein gB will be evaluated in our congenital infection model. The hypothesis we will test is that recombinant gB will provide protection against congenital CMV, but in an adjuvant-dependent fashion. We will directly evaluate whether a more potent immunomodulatory agent, monophosphoryl lipid A (MPL), provides a better adjuvant effect than alum-based adjuvants alone. These will be the first assessments of adjuvants relevant to human clinical use in the GPCMV model. In the second specific aim, we will test the role of a cytotoxic-T-lymphocyte (CTL) target, UL83, as a vectored vaccine expressed in vaccinia. We hypothesize that vaccination against this CTL target will provide protection against congenital CMV disease. This will be the first assessment of a vaccine which elicits only cell-mediated responses in a model of congenital infection. Finally, in the third specific aim we will conduct the first studies to test the protective efficacy of DNA vaccines for protection against congenital CMV infection. We hypothesize that DNA vaccines which target both gB and UL83 will provide efficacy against disease in our model. Information obtained from these studies will be relevant to ongoing studies of HCMV vaccines, and will help to prioritize which vaccine strategies should be pursued in further clinical trials for the prevention of congenital cytomegalovirus infection.